THE EARTHQUAKE ADVANCE THEORY 177
glacier causes great extension of the glacier end. If we may assume this glacier to show £ normal type of notable advance due to increase in snowfall, which might be compared with the Yakutat Bay glaciers, there is one very noteworthy difference. While the Yakutat Bay advance began and ended in a period of ten months, the Vernagt, in its last advance, moved forward much more slowly and through a much longer period of time. It was, in other words, less spasmodic. This, however, is not a vital point, for it is conceivable that with still greater increase in snowfall the advance might have been more spasmodic, as it apparently was in earlier periods when its terminus was pushed much farther down the valley than in 1899-1900.
Turning to the Yakutat Bay region, and selecting one of the glaciers there for comparison, we might almost equally well take either the Atrevida or the Galiano Glacier, but the latter will serve our purpose best because it is the smaller and therefore more nearly •comparable with the Vernagt. Its advance was great, spasmodic, and effective. It •differs from the Vernagt in having no great snowfield reservoir with which to intensify the effect of increased snowfall. To account for the advance, therefore, one would need to assume a far greater increase in snowfall than was required in the Vernagt, and exactly the same statements apply to the Atrevida Glacier. How much greater the snowfall would need to be we do not know, but surely to account for such changes as were observed would require a vast addition to the upper glacier, perhaps even more than the .greatest recorded variation in precipitation in any of the'Alaskan records, continued through several years.
Although such increase does not seem to us at all probable, we have to admit the possibility of it in view of the records of Alaskan precipitation. However, assuming such increase in precipitation, and further assuming that it was localized in the Yakutat Bay region, and that it was competent to account for the advance of Atrevida and Galiano Glaciers, we are at once confronted by the very serious difficulty that Black <jlacier, heading in the same mountains as the Galiano, and only five or six miles away, has not responded. Its failure to advance as a result of snow supply added by earthquake shaking is explicable, but that it should not have responded to such a great increase of snowfall as would be required to account for the advance of Galiano and Atrevida Glaciers is inexplicable. The Black Glacier resembles these two glaciers in being of small size, in having no enlarged reservoir, in receiving its supply from snowfields and steeply perched tributaries, and in passing down a short, deep mountain valley enclosed between precipitous walls. It should show at least some response to greatly increased snowfall; -it might fail to show response to earthquake shaking.
With these facts in mind we cannot retain the climatic hypothesis, even though granting the maximum variation in precipitation that anyone could possibly claim. But -aside from this, the hypothesis as applied to this region is weak in other respects. It must assume a localized increase in snowfall of great amount, without any explanation ^either of its localization or of its cause, and without any proof of its occurrence. In this respect the climatic hypothesis compares unfavorably with the earthquake hypothesis, •for by that we know why there was localization of effect, we know that vigorous and long continued earthquake shaking occurred, and we know that vast quantities of snow find ice were added to the glacier reservoirs. The climatic hypothesis is difficult to accept Also because of the unusual, and in fact unparalleled, phenomena which it would be •called upon to explain. While it is true that the Vernagt Glacier has advanced in a 12